[FEA]: BlockRunLengthEncode

### Is this a duplicate?

- [x] I confirmed there appear to be no [duplicate issues](https://github.com/NVIDIA/cccl/issues) for this request and that I agree to the [Code of Conduct](CODE_OF_CONDUCT.md)

### Area

CUB

### Is your feature request related to a problem? Please describe.

Hi, I've got a few applications that would benefit from the ability to do Block-level run length encoding. It looks like CUB already supports run length decoding, but the inverse is also frequently useful!

### Describe the solution you'd like

A new capability, BlockRunLengthEncode with analogous interface to the existing BlockRunLengthDecode.

### Describe alternatives you've considered

Currently, I've been mixing several of the existing block level primitives (sort + discontinuity + scan) to implement this encoding, but it's unclear to me if this is a performant option.

### Additional context

_No response_